Enter opens the highlighted result. Esc closes.

GPT-5.1 vs Llama 3.1 70B Instruct

GPT-5.1 leads Llama 3.1 70B Instruct on the Artificial Analysis Intelligence Index, 24.7 to 6.6, and Llama 3.1 70B Instruct costs 17.9x less per output token ($0.56 vs $10.00 per million). Updated October 21, 2026; 43 benchmarks have a result for both.

Replace GPT-5.1 Replace Llama 3.1 70B Instruct
GPT-5.1OpenAI
Coding
40.4
Overall
55.0
Output $/M
$10.00
43
Llama 3.1 70B InstructMeta Llama
Coding
No data
Overall
24.3
Output $/M
$0.56
0
vs boards won of 43 shared

GPT-5.1 or Llama 3.1 70B Instruct: which to pick

Only GPT-5.1 has a coding score, 40.4 on the Vibe Coding Index; Llama 3.1 70B Instruct has none, so the two cannot be ranked on coding.

Pick Llama 3.1 70B Instruct for output tokens 17.9x cheaper ($0.56 vs $10.00 per million).

Model AGPT-5.1OpenAI40.4/1003 independent benchmarks$1.02 per solved taskModel BLlama 3.1 70B InstructMeta LlamaNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
43
GPT-5.1 only
58
Llama 3.1 70B Instruct only
4
Like-for-like categories
7 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
GPT-5.1 40.4 from 3 benchmarks, Llama 3.1 70B Instruct not benchmarked.
Widest category
Agentic work, 66.7 points apart on average over 1 shared benchmark.
Widest single gap
AA Math Index: GPT-5.1 94.0 against Llama 3.1 70B Instruct 4.0, 90.0 points apart on a 0 to 100 scale.
Output price
GPT-5.1 $10.00 against Llama 3.1 70B Instruct $0.56 per million output tokens, 17.9x apart.
Context window
GPT-5.1 takes 400K tokens, Llama 3.1 70B Instruct 131K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryGPT-5.1BarsLlama 3.1 70B InstructTies
Coding50
Intelligence20
Agentic work10
Reasoning60
Math50
Knowledge50
General190
All shared boards430
CodingLike-for-likeSix of these boards feed the Vibe Coding Index50
Coding benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
SWE-bench Verified66.0%Not measuredOnly GPT-5.1 measured
LMArena WebDev1396Not measuredOnly GPT-5.1 measured
AA Coding Index49.4Not measuredOnly GPT-5.1 measured
LiveCodeBench (Vals AI)86.5%Not measuredOnly GPT-5.1 measured
SWE-bench Verified (Vals AI)69.8%Not measuredOnly GPT-5.1 measured
Vibe Code Bench (Vals AI)24.6%Not measuredOnly GPT-5.1 measured
LiveCodeBench (Artificial Analysis run)86.8%23.2%GPT-5.1
Terminal-Bench Hard (Artificial Analysis run)45.5%3.0%GPT-5.1
Terminal-Bench47.6%Not measuredOnly GPT-5.1 measured
GSO-Bench13.7%Not measuredOnly GPT-5.1 measured
WeirdML60.8%9.0%GPT-5.1
LMArena Text: coding14531260GPT-5.1
LMArena Text: software and IT14541260GPT-5.1
LMArena WebDev: React1307Not measuredOnly GPT-5.1 measured
LMArena WebDev: HTML1405Not measuredOnly GPT-5.1 measured
LMArena WebDev: from an image1421Not measuredOnly GPT-5.1 measured
Neither model has a result on the other 23 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
AA Intelligence Index24.76.6GPT-5.1
Epoch Capabilities Index149.8125.9GPT-5.1
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
τ²-Bench Telecom81.9%15.2%GPT-5.1
BALROGNot measured27.9%Only Llama 3.1 70B Instruct measured
DeepResearch Bench42.8%Not measuredOnly GPT-5.1 measured
TheAgentCompanyNot measured6.9%Only Llama 3.1 70B Instruct measured
Vending-Bench 2$1,473Not measuredOnly GPT-5.1 measured
Neither model has a result on the other 36 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index60
Reasoning benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
GPQA Diamond86.6%Not measuredOnly GPT-5.1 measured
MMMU Pro (Vals AI)83.2%Not measuredOnly GPT-5.1 measured
GPQA Diamond (Artificial Analysis run)87.3%40.9%GPT-5.1
Humanity's Last Exam (Artificial Analysis run)28.5%4.5%GPT-5.1
GPQA Diamond (Epoch AI run)87.6%44.2%GPT-5.1
ARC-AGI72.8%Not measuredOnly GPT-5.1 measured
ARC-AGI-217.6%Not measuredOnly GPT-5.1 measured
SimpleBench53.2%Not measuredOnly GPT-5.1 measured
Chess puzzles42.0%Not measuredOnly GPT-5.1 measured
Mystery game puzzles19.0%Not measuredOnly GPT-5.1 measured
VPCT58.7%Not measuredOnly GPT-5.1 measured
LMArena Text: hard prompts14571241GPT-5.1
LMArena Text: expert14701209GPT-5.1
LMArena Text: hard prompts in English14621261GPT-5.1
LMArena Vision: diagrams1264Not measuredOnly GPT-5.1 measured
LMArena Vision: homework1272Not measuredOnly GPT-5.1 measured
Neither model has a result on the other 6 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index50
Math benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
AIME93.3%Not measuredOnly GPT-5.1 measured
MGSM (Vals AI)93.0%Not measuredOnly GPT-5.1 measured
AA Math Index94.04.0GPT-5.1
AIME 2025 (Artificial Analysis run)94.0%4.0%GPT-5.1
OTIS Mock AIME88.6%3.6%GPT-5.1
MATH level 5Not measured36.7%Only Llama 3.1 70B Instruct measured
LMArena Text: math14461252GPT-5.1
LMArena Text: mathematical work14491253GPT-5.1
Neither model has a result on the other 5 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index50
Knowledge benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
MMLU Pro86.4%Not measuredOnly GPT-5.1 measured
LegalBench (Vals AI)85.7%Not measuredOnly GPT-5.1 measured
CaseLaw (Vals AI)73.4%Not measuredOnly GPT-5.1 measured
CorpFin (Vals AI)63.8%Not measuredOnly GPT-5.1 measured
TaxEval (Vals AI)74.9%Not measuredOnly GPT-5.1 measured
MortgageTax (Vals AI)61.4%Not measuredOnly GPT-5.1 measured
MedQA (Vals AI)96.4%Not measuredOnly GPT-5.1 measured
MedCode (Vals AI)52.7%Not measuredOnly GPT-5.1 measured
MMLU Pro (Artificial Analysis run)87.0%67.6%GPT-5.1
SimpleQA Verified48.0%Not measuredOnly GPT-5.1 measured
MMLUNot measured80.1%Only Llama 3.1 70B Instruct measured
LMArena Text: science14591276GPT-5.1
LMArena Text: medicine and healthcare14561257GPT-5.1
LMArena Text: legal and government14591283GPT-5.1
LMArena Text: business and finance14311235GPT-5.1
FACTS Benchmark Suite45.0%Not measuredOnly GPT-5.1 measured
Neither model has a result on the other 2 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
Web design benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
Design Arena: websites1187Not measuredOnly GPT-5.1 measured
Design Arena: UI components1166Not measuredOnly GPT-5.1 measured
Design Arena: game dev1179Not measuredOnly GPT-5.1 measured
Design Arena: data viz1204Not measuredOnly GPT-5.1 measured
Design Arena: 3D1061Not measuredOnly GPT-5.1 measured
Design Arena: all code categories1171Not measuredOnly GPT-5.1 measured
Design Arena: SVG1149Not measuredOnly GPT-5.1 measured
Design Arena: ASCII art1119Not measuredOnly GPT-5.1 measured
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index190
General benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
SAGE (Vals AI)43.2%Not measuredOnly GPT-5.1 measured
MedScribe (Vals AI)88.1%Not measuredOnly GPT-5.1 measured
IFBench72.9%34.4%GPT-5.1
Long context reasoning (AA-LCR)80.0%Not measuredOnly GPT-5.1 measured
DTBench90.1%60.0%GPT-5.1
LMCA43.9%14.8%GPT-5.1
CL-bench23.7%Not measuredOnly GPT-5.1 measured
CL-bench Life17.3%Not measuredOnly GPT-5.1 measured
LMArena Text14431261GPT-5.1
LMArena Text: instruction following14431231GPT-5.1
LMArena Text: creative writing14271232GPT-5.1
LMArena Vision1250Not measuredOnly GPT-5.1 measured
LMArena Text: English14511294GPT-5.1
LMArena Text: non-English14311219GPT-5.1
LMArena Text: Chinese14951215GPT-5.1
LMArena Text: German14381222GPT-5.1
LMArena Text: Japanese14531132GPT-5.1
LMArena Text: Korean14011140GPT-5.1
LMArena Text: French14501262GPT-5.1
LMArena Text: Spanish14331253GPT-5.1
LMArena Text: Russian14351234GPT-5.1
LMArena Text: Polish1440Not measuredOnly GPT-5.1 measured
LMArena Text: multi-turn14501256GPT-5.1
LMArena Text: longer queries14481241GPT-5.1
LMArena Text: writing and language14281236GPT-5.1
LMArena Text: entertainment and media14151230GPT-5.1
LMArena Vision: English1252Not measuredOnly GPT-5.1 measured
LMArena Vision: Chinese1270Not measuredOnly GPT-5.1 measured
LMArena Vision: reading text1255Not measuredOnly GPT-5.1 measured
LMArena Vision: humor1263Not measuredOnly GPT-5.1 measured
LMArena Vision: captioning1252Not measuredOnly GPT-5.1 measured
LMArena Vision: recognizing things1232Not measuredOnly GPT-5.1 measured
LMArena Vision: creative writing1266Not measuredOnly GPT-5.1 measured
LMArena Document1403Not measuredOnly GPT-5.1 measured
Neither model has a result on the other 12 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.1 and Llama 3.1 70B Instruct
BenchmarkGPT-5.1Llama 3.1 70B InstructLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

GPT-5.1$0.0063Fits in one request
Llama 3.1 70B Instruct$0.0008Fits in one request

Llama 3.1 70B Instruct costs less

Repository review

50K input + 3K output tokens

GPT-5.1$0.092Fits in one request
Llama 3.1 70B Instruct$0.030Fits in one request

Llama 3.1 70B Instruct costs less

Agent loop

200K input + 10K output tokens

GPT-5.1$0.350Fits in one request
Llama 3.1 70B Instruct$0.118Exceeds the context window

Llama 3.1 70B Instruct costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of GPT-5.1 and Llama 3.1 70B Instruct
SpecificationOpenAI GPT-5.1Meta Llama 3.1 70B Instruct
MakerOpenAIMeta Llama
WeightsNo public weights listedOpen weights
ReleasedNovember 13, 2025July 23, 2024
Context window400K131K
Input $/Mtok$1.25$0.56
Output $/Mtok$10.00$0.56
Coding index40.4Not benchmarked
Independent measurements behind it30
Cost per solved task$1.02Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.1 better than Llama 3.1 70B Instruct for coding?

Only GPT-5.1 has a coding score, 40.4 on the Vibe Coding Index; Llama 3.1 70B Instruct has none, so the two cannot be ranked on coding. Pick Llama 3.1 70B Instruct for output tokens 17.9x cheaper ($0.56 vs $10.00 per million).

Which is cheaper, GPT-5.1 or Llama 3.1 70B Instruct?

GPT-5.1 costs $1.25 per million input tokens and $10.00 per million output tokens; Llama 3.1 70B Instruct costs $0.56 and $0.56. Cost per solved task is known for GPT-5.1 only.

Which has the bigger context window, GPT-5.1 or Llama 3.1 70B Instruct?

GPT-5.1, with 400K tokens against 131K.

How many benchmarks is the GPT-5.1 vs Llama 3.1 70B Instruct comparison based on?

43 benchmarks have a result for both models as of October 21, 2026. GPT-5.1's coding index rests on 3 independent benchmarks and Llama 3.1 70B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next